Study Finds Surprises Among U.S. Cities Most Vulnerable to Disasters
As part of their work with START, researchers Susan Cutter, Walter Piegorsch, and Frank Hardisty conducted an analysis exploring the degree to which major urban areas in the United States are vulnerable to terrorism. The Washington Post reported on this research in its story "Terrorism Study Drops a Bombshell on Boise". In response to this story in the Washington Post, the research team provided additional information to clarify the nature of their work and its findings: 1. What is meant by vulnerability and how is this different from threat assessment or target assessment? The paper examines those pre-existing and past conditions/characteristics of people and places that influence an urban area's potential for harm from hazards and its ability to recover from hazards. It is not a threat assessment and does not examine the availability, presence or absence of terrorist targets. To suggest that Boise has more terrorist targets than other places is an incorrect interpretation of the research 2. What is "benchmark analysis"? Benchmark analysis is a quantitative approach risk assessors use to separate someone's or something's risk/vulnerability to an adverse outcome. The "something or someone" can be a person, ecosystem, community, etc., and the "adverse outcome" could be cancer, pollution damage, a natural hazard, terrorist casualty, etc. The method specifies a "benchmark" separator that distinguishes between "acceptable" risk/vulnerability and "unacceptable/adverse" risk/vulnerability. The corresponding benchmark point of the cancer dose, pollutant concentration, vulnerability index, etc., is then determined from existing data via advanced statistical calculations. Units that exceed the benchmark point are considered to be at elevated risk/vulnerability for the adverse outcome being studied, calling for greater monitoring, resource (re)allocation, and/or management, in order to avoid or alleviate the possible adverse event. 3. How was this benchmark approach used to characterize U.S. cities' vulnerability to terrorism? We viewed terrorist vulnerability in two ways: (i) vulnerability to occurrence of terrorism, and (ii) vulnerability to casualties from a terrorist incident. These were based on empirical data on actual terrorist incidents derived from the Global Terrorism Database at the START Center. In each case, we specified the benchmark separation point at 50%, giving a simple "median" separator of vulnerability. We then further classified communities into three levels: "green" if both benchmark indices (terrorist occurrence and terrorist casualties) were below their respective median separators, "yellow" if one was below and the other was above, and "red" if both were above. 4. Is an urban center in the "green" or "yellow" tier at lower risk of terrorist attack? No. As noted above, our benchmarking was performed on a vulnerability scale; what is being measured is a community's ability to respond to the occurrence and consequences of adverse natural or artificial events (e.g., in our study, terrorist events). The color coding was intended only to highlight the sometimes-surprising differences in vulnerability we discovered among these 132 U.S. cities. Even the least vulnerable community in our study (Juneau, AK) is at some risk for occurrence of and for casualties from a terrorist attack. Our results are intended to bring awareness of urban vulnerabilities to terrorism, and to encourage further coordination of and resource allocations towards urban emergency management/first-response capability; no community should feel complacent about their vulnerability to terrorism. 5. Why Boise, and not Los Angeles or other western cities? This was a surprise to us. Part of this is related to the diversity of natural hazards between the two and in the overall frequency of specific events. The underlying data used do not consider the risk or probability events, but rather events that have occurred (1960-2006) for which there was a monetary loss. In many ways, it under-reports events like earthquakes which are infrequent; and over reports weather-related events. It also measures losses as a percentage of county gross domestic product. This means that a $1 million loss in Los Angeles with a very high GDP has less impact, relatively speaking, than the same $1 million loss in Boise (or Ada county) with a lower GDP. This (more diversity in hazards, more events, higher percentage of relative loss) helps explain why Boise is higher than Los Angeles on this particular component. 6. What would you tell the residents of the Boise metropolitan area based on your study? First, they are not a "targeted area", but they have experienced a terrorist incident and are susceptible to them in the future. Second, the metro region has average levels of social vulnerability suggesting that the resident population has the social networks and capacity to respond to disasters including terrorist threats. The infrastructure component was low based on lower urban densities, fewer landmark buildings, and this bodes well for the metro region. The higher values on the hazards score are reflective of the experience in the metro area with natural hazards and their impacts. Instead of looking at this negatively, I would look at it positively and suggest that Boise has had more experience with emergency response and recovery and this will come in handy should they experience another natural hazard or terrorist event. Note: There was an error in the Washington Post story on the funding. $4 million was provided to the START Center headquartered at the University of Maryland, not the University of South Carolina. South Carolina is a partner in the national center.